ESG equities and Bitcoin: responsible investment and risk management perspective

Abstract

Purpose

While an increasing number of investors value socially responsible investment practices, Bitcoin has faced criticism for its carbon footprint resulting from excessive mining power consumption. By examining Bitcoin’s interconnectedness with environmental, social and governance (ESG) equities, this study aims to construct a socially responsible investment strategy for cypto investors.

Design/methodology/approach

This study uses wavelet analysis and a time-varying parameter vector autoregressive (TVP-VAR) model to uncover the interdependence between ESG equities and Bitcoin. This study computes the optimal ratio, showing that Bitcoin significantly reduces portfolio risk when combined with green stocks.

Findings

The results show that co-movements between green stocks and Bitcoin are low, indicating that they are suitable combinations for portfolio diversification. From an environmental perspective, this investment strategy offers a theoretical solution to mitigate the negative impacts associated with Bitcoin mining. It aims to address the dilemma faced by sustainability-conscious investors, who must navigate the economic payoff of Bitcoin against their commitment to green investment principles.

Practical implications

The findings can provide valuable insights for policymakers seeking to develop strategies that promote sustainable investments among crypto investors.

Originality/value

Research on ethical investment practices in the cryptocurrency market remains in the early stages of development. Ethical investors can benefit from including Bitcoin in their ESG equity portfolios.

Citation

Kakinuma, Y. (2024), "ESG equities and Bitcoin: responsible investment and risk management perspective", International Journal of Ethics and Systems, Vol. 40 No. 4, pp. 759-775. https://doi.org/10.1108/IJOES-03-2023-0049

1. Introduction

Cryptocurrency has established itself as an alternative investment, owing to the growing popularity of decentralized features that are free from government interference. Blockchain, a revolutionary innovation equivalent to the internet, enables decentralized systems with a distributed digital ledger. Implementing the blockchain technology the most successfully, Bitcoin is the most dominant cryptocurrency, and its unique pricing mechanism contributes to a risk management strategy as a diversifier, hedge or safe haven for traditional financial assets (Bouri et al., 2017; Urquhart and Zhang, 2019; Conlon and McGee, 2020; Huang et al., 2021). With these unique benefits, Bitcoin is becoming an increasingly standard part of portfolios for both institutional and retail investors (Nedved and Kristoufek, 2023). Nonetheless, Bitcoin is subject to criticism for its excessive energy consumption because its proof-of-work mining demands significant computational resources. This translates to a considerable carbon footprint. Bitcoin’s electricity consumption and resulting carbon emissions are so large that they are comparable to entire countries (Stoll et al., 2019). Truby (2018) and Delgado-Mohatar et al. (2019) caution that the Bitcoin network operates with inefficient use of scarce resources. Another critique of cryptocurrencies is that they facilitate financial crimes such as terrorist financing, money laundering and corruption (Teichmann and Falker, 2021). The anonymity and decentralized peer-to-peer transmission system enabled by blockchain technology has created an opportunity to carry out the illegal financial activities, and Bitcoin dominates cryptocurrency-related crimes (Kethineni and Cao, 2020).

Socially responsible investment (SRI) considers nonfinancial criteria, such as environmental and social impacts. In principle, SRI avoids assets that have poor social and environmental records. However, we propose that ethical crypto investors can take advantage of environmental, social and governance (ESG) compliance assets to counter the negative aspects of Bitcoin. Promoted by governments and policymakers, ESG or sustainable investing has gained popularity, and its economic significance, such as risk reduction (Kim et al., 2014; Albuquerque et al., 2019), has driven investors’ demand for ESG assets. Among sustainability-linked assets, green bonds have been extensively studied as possible diversifiers of cryptocurrencies (Hung, 2021; Le et al., 2021; Yadav et al., 2022). We focus on green equities for SRI-principled crypto investors because studies on the link between ESG equities and the crypto market are scarce. We explore an alternative sustainable investing scheme following the findings of Gabriel et al. (2022) that low-carbon equity indices do not behave like conventional indices.

Our analysis focuses on ESG equities in Asia and Bitcoin. Feyen et al. (2022) report that Asia is one of the regions that leads crypto transaction volumes. In addition, Japan, China and South Korea are identified as fast-growing blockchain hubs based on the concentration of tech firms with R&D commitment to blockchain projects, the number of blockchain patents obtained and their digital and regulatory infrastructure (Wang et al., 2019a, 2019b). More importantly, China accounts for more than 75% of the world’s Bitcoin mining (Jiang et al., 2021), which consumes an enormous amount of energy and results in excessive carbon emissions. Thus, the implications of this study are particularly important for sustainability-conscious investors in this promising region of blockchain and cryptocurrency development. We propose an investment principle with green equities and cryptocurrency, and the benefits are twofold: investment risk reduction and alleviation of the adverse environmental impacts of blockchain mining.

This study makes a contribution to the existing literature on the relationship between fintech development and the environment. While technological advancement has often seen as a double-edged sword in terms of its environmental impact (Tao et al., 2022), recent reviews of relevant articles have demonstrated that innovation and sustainability serve as the fundamental driving forces behind the financial industry (Puschmann et al., 2020). Digital currencies using cryptographic blockchain authentication for secure transactions are transformative technologies that do not depend on traditional banking or government regulatory systems. However, the excessive energy used for cryptocurrency mining causes environmental and social damage, and leads to negative net social benefits (Goodkind et al., 2020). Amidst growing concerns for climate impacts, recent literature suggests fintech development can have positive impact on environment protection initiatives. For example, using fintech startups (Croutzet and Dabbous, 2021), fintech usage (Muganyi et al., 2021) and market-cap-weighted fintech company index (Nenavath, 2022) as a proxy for fintech development, the literature shows that advancement in financial technology reduces carbon emissions and promotes renewable energy. To further investigate the nexus between technological development and environmental impact, we follow Goodell et al. (2022) and examine the connectedness between the two factors. Our contribution lies in highlighting the potential benefits for ethical investors who incorporate Bitcoin into ESG equity portfolios. We show that the interdependence between ESG equities and Bitcoin is low, and that Bitcoin offers significant risk reduction when combined with green stocks. From an environmental perspective, this investment strategy reconciles the economic profits of Bitcoin investments with society’s sustainability.

The remainder of this paper is organized as follows: Section 2 reviews the related literature. Section 3 describes the data set and methodological approach. Section 4 presents the results, and Section 5 summarizes the main findings and concludes.

2. Literature review

2.1 Connectedness between Bitcoin and traditional financial markets

Often labeled as digital gold, the idea of Bitcoin as a safe haven asset has been the focus of risk management literature. Studies have found that Bitcoin is weakly or negatively correlated with equity markets (Dyhrberg, 2016; Bouri et al., 2017; Bouri et al., 2020a, 2020b; Baur et al., 2018; Ji et al., 2018; Fang et al., 2019). The diversifying abilities of Bitcoin in equity markets can be attributed to its unique price-driving forces (Kristoufek, 2015; Bouoiyour et al., 2016), including factors such as anonymity (Möser and Böhme, 2017), hash rate (Kubal and Kristoufek, 2022), energy prices (Sarker et al., 2023) and energy consumption (Huynh et al., 2022). While some studies argue that Bitcoin is a panacea (Bouri et al., 2020a, 2020b), gold has been found to be a superior and more stable safe haven than Bitcoin (Shahzad et al., 2020; Kakinuma, 2022; Ustaoglu, 2023). More importantly, although these studies praise Bitcoin’s unique hedging abilities against equities, few mention the drawback of the excess electricity consumption required for the Bitcoin network, which raises serious environmental and ethical concerns (Jalan et al., 2023).

The diversification properties of Bitcoin have been examined in relation to various financial assets such as forex (Baumöhl, 2019), oil (Zeng et al., 2020; Lin and An, 2021), natural gas (Lin and An, 2021), real estate investment trusts (Alam et al., 2023) and bonds (Maghyereh and Abdoh, 2022). These results suggest that Bitcoin is outside the global financial system despite its growing popularity. This finding is valuable for investors and portfolio managers in integrated financial markets where diversification opportunities are becoming limited. Nonetheless, Bitcoin’s extremely high volatility is of great concern, and possible volatility spillovers to other markets threaten financial stability. Zhang et al. (2021) found a downside volatility spillover between Bitcoin and conventional assets, challenging the earlier literature’s claim that Bitcoin is isolated from the global financial markets (Bouri et al., 2020a, 2020b; Li and Huang, 2020). Zha et al. (2023) also show significant risk spillover between Bitcoin and traditional financial markets, and that the COVID-19 outbreak enhanced the interplay.

2.2 Connectedness between Bitcoin and green assets

While green bonds have received considerable attention as diversifiers in the cryptocurrency market (Karim et al., 2022; Yadav et al., 2022; Lee et al., 2023; Ul Haq et al., 2023), research on the relationship between Bitcoin and green equities is scarce in the literature. Jiang and Zhang (2023) conclude that the prices of carbon credit futures and Bitcoin can be a good hedge against each other during normal periods but not during high uncertainty such as the COVID-19 pandemic. Similarly, Anwer et al. (2023) indicate that sustainable equities and crypto assets moved together during the COVID-19 crisis and did not offer diversification benefits. In contrast, Fareed et al. (2022) recommend holding both low-carbon stocks and Bitcoin during the pandemic, as the crisis positively impacted the two assets. Thus, there is a discrepancy in the literature concerning the risk mitigation properties of digital assets and green equities. Naeem and Karim (2021) found that Bitcoin can be hedged with clean energy stocks. However, their hedging strategy involves taking a long position in Bitcoin and a short position in clean energy stocks. From the perspective of ethical investing, short selling green financial assets signify a departure from environmentally friendly investment principles. Here, we attempt to bridge this gap in the literature by analyzing a diversification strategy using Bitcoin and green equities, which also complies with a sustainable investment philosophy.

3. Data and methodology

We consider the MSCI ESG Leaders Indexes as proxies for ESG equity indices. The MSCI ESG Leaders Indexes track firms with the highest ESG rating and strong sustainability profiles in each country. We use the ESG index prices of Japan (JPN), China (CHN), Hong Kong (HKG), South Korea (KOR), Thailand (THA) and Indonesia (IDN) in our analysis. The ESG indices and BTC prices were retrieved from Datastream for the period from January 1, 2014, to November 30, 2022. Our sample period starts in 2014 because BTC’s trading volume before 2013 was too low and insufficient (Wang et al., 2019a, 2019b). Table 1 reports the summary statistics for the daily log-return series. BTC exhibits the highest mean returns with the largest volatility. Jaque-Bera test and ADF test suggest that all series are not normally distributed and stationary.

We apply the wavelet coherence approach of Grinsted et al. (2004), which allows us to identify the co-movement between ESG indices and BTC in both the time and frequency domains. Following Torrence and Compo (1998), the cross-wavelet transform of the time series x(t) and y(t) with the continuous wavelet transform (CWT) Wnx (u, s) and Wny (u, s) is defined as:

(1) Wnxy(u, s)=Wnx(u,s)∗Wny(u,s)

where u is the location and s is the scale. * represents the complex conjugate. The CWT unveils areas in the time–frequency domain where the time series exhibit high interaction.

Following Torrence and Webster (1999), we define the wavelet coherence of the two time series as follows:

(2) R2(u,s)=|S(s−1Wxy(u,s))|2S(s−1|Wx(u,s)|2)S(s−1|Wy(u,s)|2)

where S denotes the smoothing operator over time and scale. The coefficient of the squared wavelet coherence R²(u, s) ranges from 0 to 1, with 0 indicating zero co-movement between the time series and 1 indicating strong co-movement.

For an additional robustness analysis, we use the time-varying parameter vector autoregression (TVP-VAR) model of Antonakakis et al. (2020), which is an extension of an original approach by Diebold and Yilmaz (2012, 2014). The proposed model over the original model includes the elimination of an arbitrary rolling window setup, reduced sensitivity to outliers and no loss of observations in the dynamic measure calculation. We estimate a TVP-VAR model with a lag length of order one, selected using the Bayesian information criterion (BIC):

(3) Yt=ΦtYt−1+εt, εt ∼ N(0, St)

(4) vec(Φt)=vec(Φt−1)+ut,ut ∼ N(0, Rt )

where Y_t and ε_t are N × 1 dimensional vectors, Φ_t and S_t are N × N dimensional matrices, vec(Φ_t) and u_t are N² × 1 dimensional vectors and R_t is a N² × N² matrix.

Next, we calculate the H-step ahead generalized forecast error variance decomposition (GFEVD) as defined by Koop et al. (1996) and Pesaran and Shin (1998). Using the World representation theorem, equation (3) can be transformed into a TVP-vector moving average model: Yt=∑i=1pΦitYt−1+εt=∑j=0∞Ajtεt−j+εt. The GFEVD represents the forecast error variance share variable i explains variable j, or, in other words, the pairwise directional connectedness from j to i, which is computed by:

(5) Φij,tg (H)=Sii,t−1∑t=1H−1(ι ′iAiSiι j)2∑j=1k∑t=1H−1(ι iAiSiA′iι i) Φ˜ij,tg (H)=Φij,tg (H)∑j=1kΦij,tg (H)

where ∑j=1kΦ˜ij,tg (H)=1, ∑i,j=1kΦ˜ij,tg (H)=k and ι_i is a zero vector with unity on the ith position. Based on the GFEVD, and following Gabauer and Gupta (2020), we estimate the following connectedness measures:

(6) TOjt=∑i=1,i≠jkΦ˜ij,tg (H)

(7) FROMjt=∑i=1,i≠jkΦ˜ji,tg (H)

(8) NETjt=TOjt−FROMjt

(9) TCIt=k−1∑j=1kTOjt≡k−1∑j=1kFROMjt

(10) NPDCij,t=Φ˜ij,tg (H)−Φ˜ji,tg (H)

Equation (6) represents the directional spillover that variable j transmits to all other variables, defined as the total directional connectedness to others, whereas equation (7) indicates the directional spillover that variable j receives from all other variables, defined as the total directional connectedness from others. Equation (8) is the net total directional connectedness, which is the difference between total directional connectedness to others and total directional connectedness from others. The total connectedness index (TCI) is expressed by equation (9), which is the average shock that one variable has on all other variables. The TCI measures how tightly or loosely an entire network is interconnected. Finally, we calculate the net pairwise directional connectedness (NPDC) given by equation (10). A positive NPDC indicates that variable i dictates variable j, whereas a negative NPDC indicates that variable j dominates variable i.

We further construct portfolio allocation strategies with ESG indices and BTC using the dynamic conditional correlations (DCC) obtained from Engles (2002) DCC-generalized autoregressive conditional heteroskedasticity model. Following Kroner and Ng (1998), the risk-minimizing optimal portfolio weights without reducing expected returns are defined as follows:

(11) wyt= hyyt−hxythxxt−2hxyt−hyyt

where w_yt is the optimal weight of asset y, h_xy is the conditional covariance of assets x and y and h_xx and h_yy are the conditional variances of x and y. In this study, y is BTC and x is the ESG index. Because of the no-shorting constraint, the following restriction applies:

(12) wyt={0     if wyt<0wyt if 0 ≤ wyt ≤11     if wyt >1

We evaluate the effectiveness of portfolio allocation strategies by hedging effectiveness (HE):

(13) HE=Varunhedged−VarhedgedVarunhedged

where Var_unhedged denotes the variance of returns of an ESG index and Var_unhedged refers to the variance of the returns of the ESG index-BTC portfolio. We compute the statistical significance of the HE using Brown and Forsythe’s (1974) test of equality of variances.

4. Results

4.1 Wavelet coherence analysis

Figure 1 shows the wavelet coherence between the ESG indices and BTC, which captures the comovement between the two series in the time frequency space. The horizontal and vertical axes represent the time and frequency domains, respectively. The frequency domain represents investment horizons, which we define as 4–16 days for the short-term, 16–64 days for the medium-term and 64–256 days for the long-term.

The results show that regardless of country, cold colors (blue) dominate the short-term interaction between the ESG index and BTC, indicating a lack of co-movement. The short-term isolation of BTC from sustainable investments supports Goodell et al.’s (2022) findings. Red islands appear in early 2020 for all ESG indices in the mid-term horizon, representing significant coherence with BTC during the COVID-19 outbreak. The increased connectedness between the equity market and BTC during the market downturn is in line with Conlon and McGee (2020). In the long-term analysis, JPN ESG shows high co-movement with BTC during 2014–2018. This strong interaction is unique to the Japanese market and could be attributed to the closure of Mt. Gox in 2014, a Tokyo-based Bitcoin exchange platform that accounted for 70% of BTC trading volume before the closure (Decker and Wattenhofer, 2014). THA ESG has strong long-term coherence with BTC starting in late 2017.

4.2 Time-varying parameter vector autoregressive model

Table 2 reports the average connectedness estimated by the TVP-VAR. The diagonal elements show own shocks, whereas off-diagonal elements indicate inter-volatility spillovers. Our main interests are the interaction between BTC and the ESG indices. The first column of Table 2 reveals that negligible variation in BTC returns is attributable to the impact of innovation in the ESG indices, specifically, 1.51% for JPN, 0.90% for CHN, 0.58% for HKG, 1.07% for THA and 0.53% for IDN. BTC’s total directional connectedness to all ESG indices is limited to 5.67%, confirming the weak integration between BTC and financial markets. Similarly, BTC receives minimal shocks from all ESG indices, as shown in the first row of Table 2. BTC draws 1.14% spillover from JPN, 1.46% from CHN, 1.15% from HKG, 1.53% from KOR, 1.51% from THA and 0.96% from IDN. The total directional connectedness from all ESG indices to BTC is 7.76%. The TCI is 36.26%; however, most of the integration is from the interaction among the ESG indices. These results are consistent with Trabelsi (2018) and Zeng et al. (2020) who find no significant spillover effects between BTC and conventional assets.

Next, we focus on the time-varying directional contribution from BTC (all ESG indices) to all ESG indices (BTC), as plotted in Figure 2. In both directions, the spillover index remains low and flat most of the time, except at the beginning of the sample period and in early 2020, which corresponds to the hacking of Mt. Gox and the emergence of COVID-19. Nonetheless, even these spikes are moderate and short-lived. Turning our attention to time-varying net pairwise directional connectedness (see Figure 3), it is evident that shocks to any of the ESG indices do not affect BTC for most of the time period. The spillover index is mostly flat and ranges from positive to negative five for the CHN, HKG, THA and IDN ESG equity pairs. BTC is a net volatility transmitter for JPN and KOR ESG stocks in 2022, which is in line with Hasan et al. (2023), who found that Japanese ESG assets received volatility transmissions in the post-COVID period.

Overall, according to our dynamic connectedness analysis using the TVP-VAR model, BTC is largely disconnected from ESG equities over the sample period, providing valuable insights for investors seeking to develop effective diversification strategies using cryptocurrency and green assets.

4.3 Optimal portfolio allocation

Our wavelet coherence analysis and TVP-VAR model suggest that the interconnectedness between ESG indices and BTC is weak, which offers diversification opportunities. Thus, we construct portfolio allocation strategies. Following Basher and Sadorsky (2016), we forecast the conditional volatility (one period ahead) to generate a one-day-ahead optimal ratio. We use a fixed-width rolling window with daily returns to generate 50 out-of-sample optimal ratio forecasts, with the model refitted for every ten observations.

Table 3 reports the average optional ratios of BTC, w_BTC, when combined with the ESG index. The optimal ratios vary across ESG indices, ranging from 32% for CHN to 2% for THA. Hedging effectiveness is statistically significant for all ESG indices, indicating an important implication that BTC can be an effective diversification tool from the perspective of investment risk management and offsetting concerns regarding adverse environmental impacts of BTC mining. The inclusion of BTC in ESG portfolios leads to considerable risk reduction, especially for CHN (35.79%) and HKG (22.66%). The diversification benefits are limited for IDN (9.20%) and THA (4.23%); however, highly volatile BTC can provide a small but significant risk reduction for green investors.

We conduct a robustness check by changing the forecast length. Table 4 reports the estimated optimal weight and HE for forecast lengths of 100 and 200 days. These results are consistent, although we forecast the longer lengths. The optimal BTC weights in the ESG portfolio decrease slightly as we extend the forecast period, but all HE are significant. CHN receives the greatest diversification benefits from BTC. China accounts for more than three-fourths of the hashing power of the Bitcoin network (Jiang et al., 2021), leading to an excessive carbon footprint. By holding both ESG stocks and BTC, green-conscious investors in China can diversify their investment risk and counteract the environmental impact of blockchain mining.

5. Conclusion

Our investigation focused on examining the interconnectedness between Asian green equities and Bitcoin, to explore diversification opportunities for investment risk management while mitigating the environmental impacts associated with Bitcoin mining. Our wavelet analysis detects low co-movements between green financial assets and Bitcoin and supports the findings of Uddin et al. (2020) and Goodell et al. (2022). As a robustness check, we use the TVP-VAR approach to analyze the dynamic connectedness among different asset classes. Consistent with Trabelsi (2018) and Zeng et al. (2020), we found no significant spillover effects between Bitcoin and sustainable equities. We further computed the optimal Bitcoin ratio in the green portfolio to obtain a practical risk management strategy. The results reveal that holding a handful of Bitcoin effectively reduces the volatility of green stock portfolios. Among our samples, ESG equities in China benefit the most from Bitcoin allocation, indicating an important implication for sustainable investors in the country where the majority of world blockchain mining is conducted with considerable energy consumption. Our findings are robust to alternative model specifications.

We contribute to the literature by highlighting that constructing green portfolios with Bitcoin offers risk-reduction benefits while simultaneously combining financial motives with sustainable investing. While the absence of a significant link between Bitcoin and green assets may suggest that cryptocurrency market growth does not contribute to the development of sustainable investment instruments (Corbet et al., 2021), we view it as an opportunity for ethical investors to proactively create sustainable portfolios aimed at combating climate change. We support Ren and Lucey (2022) that portfolio stability and ecological protection are not necessarily mutually exclusive. This is contrary to the results of Naeem and Karim (2021), who show that green assets serve as a hedge for Bitcoin but by short-selling sustainable assets. The benefits of portfolio risk reduction at the expense of selling green assets raise concerns for socially responsible investors. Ciaian et al. (2022) point out a strong association between investors’ ESG preferences and cryptocurrency ownership; some crypto investors value nonpecuniary profits, such as ethics and morals. The combination of green stocks and Bitcoin provides a remedy for the sustainability-conscious investors’ dilemma of the economic payoff of Bitcoin versus adherence to the green investment philosophy. Our results also extend those of Díaz et al. (2022), who show that SRI mitigates the risks of investing in Bitcoin using a single clean energy ETF as a proxy for green equity. Instead of relying on a single asset, we used ESG equity indices from different Asian countries. We confirm that the diversification benefits of ESG equities for crypto investors are not region-sensitive, as all Asian ESG indices included in this study benefited significantly from the inclusion of Bitcoin. Finally, our findings are consistent with Goodell et al. (2022) in demonstrating that Bitcoin and green assets effectively diversify each other, offering a tandem environmental offset.

Our results offer several key policy implications. First, crypto investors should be reminded that they can follow SRI principles and contribute to sustainability through ESG investments. Simultaneously, policymakers should promote sustainable investing, particularly by targeting crypto investors and emphasizing both economic and social benefits. Often seen as a speculative asset rather than a long-term wealth creation vehicle, cryptocurrency investment is associated with an immediate payoff without considering its impact on society and the environment. In this sense, regulators and academics play an important role in equipping investors with broader knowledge and raising their awareness of climate change arising from their investment choices. Second, policymakers can simultaneously explore the impact of an environmental tax on cryptocurrency transactions and provide tax incentives for ESG investments. This may help promote sustainable investing among crypto investors, who can benefit from risk reduction in their portfolios. Third, regulators should establish standards for measuring and reporting the environmental impacts of Bitcoin mining. While there are several global initiatives aimed at promoting standardized ESG reporting practices, such as the Global Reporting Initiative and the Sustainability Accounting Standards Board, there are no universally accepted environmental impact reporting standards for Bitcoin miners. Although cryptocurrencies backed by blockchain technology have several practical advantages and the potential to transform the financial industry, negative environmental impacts, such as the excessive carbon footprint caused by Bitcoin mining, need to be addressed. The establishment of reporting standards is a step toward greater transparency and accountability in the blockchain industry. Fourth, Bitcoin miners should be encouraged to use renewable energy sources. Miners in Asia rely primarily on electricity generated from coal and natural gas (Yan et al., 2022). Instead, Bitcoin miners could consider transitioning to more sustainable energy sources such as solar and wind power. As green finance continues to grow, cryptocurrencies that prioritize eco-friendliness have greater opportunities for investment and growth.

This study invites further research into the dynamics between sustainable investments and cryptocurrencies. We constructed a portfolio with ESG equities and Bitcoin; however, other combinations of green and digital assets can be explored. Recent studies suggest that investor attention affects green equity (Gao et al., 2023) and cryptocurrencies (Smales, 2022). Thus, the connectedness of the two asset classes can be analyzed from the perspective of behavioral finance.

Figures

Figure 1.

Wavelet coherence between ESG index and Bitcoin

Figure 2.

Dynamic directional spillovers between BTC and ESG equities

Figure 3.

Dynamic net pairwise directional spillovers between BTC and ESG equities

Table 1.

Summary statistics of daily return

	JPN	CHN	HKG	KOR	THA	IDN	BTC
Minimum	−0.07783	−0.09548	−0.06898	−0.11345	−0.10787	−0.08541	−0.49397
Maximum	0.07735	0.15552	0.06621	0.09317	0.07333	0.15198	0.23839
Mean	0.00018	0.00009	0.00009	0.00011	0.00007	0.00023	0.00133
Stdv	0.01162	0.01623	0.01078	0.01140	0.00998	0.01318	0.04620
Jaque–Bera test	2,576*	4,195*	2,710*	9,811*	34,209*	12,659*	8,815*
ADF test	−13.539*	−13.191*	−13.464*	−12.52*	−12.391*	−14.482	−11.487*
Observations	2,326	2,326	2,326	2,326	2,326	2,326	2,326

Note:

*Represents significance at the 1% level

Source: Author’s own work

Table 2.

Averaged joint connectedness

	BTC	JPN	CHN	HKG	KOR	THA	IDN	FROM others
BTC	92.24	1.14	1.46	1.15	1.53	1.51	0.96	7.76
JPN	1.51	64.10	7.46	7.19	12.25	4.65	2.84	35.90
CHN	0.90	7.58	41.03	26.32	12.18	6.67	5.32	58.97
HKG	0.58	7.65	28.03	41.59	11.75	6.45	3.95	58.41
KOR	1.08	11.38	10.72	10.03	56.14	5.71	4.94	43.86
THA	1.07	3.94	5.80	5.36	5.65	73.56	4.62	26.44
IDN	0.53	2.74	5.16	3.45	5.50	5.08	77.53	22.47
TO others	5.67	34.43	58.63	53.51	48.86	30.06	22.63	253.80
NET spillovers	−2.08	−1.47	−0.34	−4.89	5.01	3.62	0.16	TCI 36.26

Notes:

The table reports the total, directional and pairwise spillovers estimated by a TVP-VAR model with lag length of order 1 and a 10-step-ahead generalized forecast error variance decomposition (GFEVD). The lag length is selected based on the BIC. The sample period is from January 1, 2014, to November 30, 2022. FROM directional values denote spillovers from all variables j to variable i. TO directional values denote spillovers from to alle i to all variables j. NET spillovers are TO minus FROM. Total spillover index (TCI) measures the overall connectedness of all variables in the model

Source: Author’s own work

Table 3.

Optimal ratio and hedging effectiveness

Forecast length	50
ESG index	Optimal weight (wBTC)	Minimum	Maximum	Hedging effectiveness
JPN	0.0799	0.0116	0.1901	0.1119*
CHN	0.3217	0.1071	0.5753	0.3579***
HKG	0.2108	0.0652	0.4177	0.2266***
KOR	0.1342	0.0430	0.2106	0.1112***
THA	0.0220	0.0046	0.0447	0.0423*
IDN	0.0692	0.0147	0.1460	0.0920***

Notes:

Forecasts of one-step ahead optimal ratios from a fixed-width rolling window.

Models are refitted every ten observations. The Brown and Forsythe (1974) test estimates the significance of hedging effectiveness.

*, ** and ***represent significance at the 10, 5 and 1% levels, respectively

Source: Author’s own work

Table 4.

Optimal ratio and hedging effectiveness – with different forecast lengths

Forecast length	100
ESG index	Optimal weight (wBTC)	Minimum	Maximum	Hedging effectiveness
JPN	0.0702	0.0192	0.1858	0.0859**
CHN	0.2244	0.0355	0.5881	0.2701***
HKG	0.1386	0.0140	0.4234	0.1599**
KOR	0.1007	0.0243	0.2144	0.1118***
THA	0.0174	0.0000	0.0445	0.0353*
IDN	0.0595	0.0139	0.1482	0.0815***
Forecast length	200
JPN	0.0652	0.0034	0.1855	0.0791**
CHN	0.2335	0.0070	0.6457	0.2823***
HKG	0.1076	0.0000	0.4215	0.1252***
KOR	0.0813	0.0000	0.2146	0.1016***
THA	0.0166	0.0000	0.0451	0.0339*
IDN	0.0594	0.0085	0.1492	0.0820*

Notes:

Forecasts of one-step ahead optimal ratios from fixed-width rolling window.

Models are refitted every 10 observations.

*, ** and ***represent significance at the 10, 5 and 1% levels, respectively

Source: Author’s own work