← All podcasts
Issue 1/2023 · Pénzügy, adózás, számvitel

Factors influencing the returns of cryptocurrencies under varying market conditions

Kovács Edina

Published: 2023 · Forum on Economics and Business, Issue 1/2023 · Pages: 20–34

„Mitől mozog a bitcoin árfolyama?”

Listen to the podcast

Conversation · 8 min
HU EN Podcast language: Hungarian · EN: English version not available yet
0:00-8:23

Published: 2023 · Forum on Economics and Business, Issue 1/2023 · Pages: 20–34

The full article HU EN

OJS ↗ Full issue (PDF) ↗

This English text is a machine translation of the original Hungarian article, made with a local AI model. The original is authoritative: read the original (Hungarian)

Közgazdász Fórum · Issue 1/2023 · 20–34

Factors influencing the returns of cryptocurrencies under varying market conditions

Abstract. In recent years, financial professionals, the media, regulatory authorities, and investors have demonstrated increasing interest in cryptocurrencies, particularly Bitcoin. This is unsurprising, given that the value of Bitcoin rose from $13 in 2013 to $42,000 by the end of 2023. This research examines the sensitivity of the returns of Bitcoin, Ethereum, and Binance to changes in factors such as gold and equity market returns, the volatility index, the Amihud illiquidity ratio, and the volume of Bitcoin-related searches as measured by Google Trends, during the period between 1 January 2018 and 31 December 2023. Throughout the study, I applied the ordinary least squares method to three distinct sub-periods, taking into account the structural breakpoints identified by the Bai–Perron test. Based on the results, the investigated factors exert varying degrees of influence across the different periods. In the first period, when the crypto market was relatively stable, the returns of the investigated cryptocurrencies were primarily influenced by gold and the volatility index. In the second period, characterised by market turbulence, the volume of searches for Bitcoin and equity market returns were the most significant factors, whereas in the third period, during the market downturn, equity market returns became the most decisive.

Keywords: bitcoin, cryptocurrencies, structural breaks

Introduction

Since the emergence of Bitcoin in 2009, numerous cryptocurrencies have appeared over the past decade, such as Ethereum (2015), Tether (2015), and Binance (2017), followed by several thousand others in recent years. Concurrently, cryptocurrency markets have garnered increasing attention from financial professionals, the media, regulatory authorities, investors, and researchers. Numerous empirical studies have examined the factors influencing cryptocurrency returns. The literature has investigated the impact of stock market returns and prices (Malladi et al. 2021; Nguyen 2022; Ahmed et al. 2023), gold (Panagiotidis et al. 2018; Jareño et al. 2020; Huynh et al. 2020), economic and political instability (Shaikh 2020), oil price changes (Salisu et al. 2023), regulatory news (Auer and Claessens 2018), and the volume of Google searches (Kristoufek 2013).

Among cryptocurrencies, Bitcoin has been the most extensively researched. Over the past ten years, the price of Bitcoin has risen from $13 in 2013 to approximately $42,000 by the end of 2023 (Coinmarketcap 2023). This period included sub-periods of high volatility; for instance, during the Covid-19 pandemic in 2020, the Bitcoin exchange rate increased by 300% from July to December (Bloomberg 2021). In April 2021, the Bitcoin price approached $60,000, halved by mid-July 2021, and subsequently reached its highest price within the studied period ($67,549) in November 2021.

Such significant price movements are uncharacteristic of traditional currencies or stock market prices, suggesting that cryptocurrency prices may be influenced by factors distinct from those affecting traditional currencies. Furthermore, the impact of these factors may vary across different periods; for example, during a high-volatility period, the volume of Google searches may be a more significant factor than changes in the price of gold. Previous studies (e.g., Auer–Claessens 2018; Baker et al. 2016; Jareño et al. 2020) did not account for the variations in the impact of the examined factors between different time periods. Therefore, this study examines the impact of individual factors on various cryptocurrencies on a period-by-period basis between 1 January 2018 and 31 December 2023.

The remainder of this article is structured as follows: the first chapter presents the literature regarding the factors influencing cryptocurrency returns. The second chapter describes the data and the methodological approach. The third chapter presents and analyses the empirical results.

The final chapter summarises the main conclusions of the research.

Literature Review and Theoretical Framework

The first step in identifying the factors influencing cryptocurrency returns was to identify relevant academic articles in order to obtain a comprehensive overview.

From the scientific research found in the Scopus and Google Scholar databases, I selected the most relevant studies based on their titles and keywords. I conducted a thorough study of those articles published based on the terms "factors influencing bitcoin returns", "cryptocurrency exchange rate", "cryptocurrency return", "factors influencing cryptocurrency returns", and related expressions.

Following the first step of the literature review, it can be established that previous studies have examined stock market returns and prices (Malladi et al. 2021; Nguyen 2022; Ahmed et al. 2023), gold (Panagiotidis et al. 2018; Jareño et al. 2020; Huynh et al. 2020), the number of Google searches (Kristoufek 2013; Chang et al. 2021), market uncertainty (Chang et al. 2021), illiquidity (Eross et

There is a causal relationship between bitcoin returns and Google search volumes in both the short and long term. Unemployment news regarding the United States has a significant impact on bitcoin returns. News related to GDP and the consumer price index are not significant.

S&P500 index, VIX, number of Google searches

Table or figure – see the PDF version of the issue for its exact content.

The rise in the gold price relative to the platinum price increases the bitcoin return.

There is a significant negative relationship between returns and volatility; however, this relationship is not return- or volatility-driven, but rather a bidirectional relationship.

Table or figure – see the PDF version of the issue for its exact content.

Ethereum and Binance yields have a significant positive effect on Bitcoin yields.

Table or figure – see the PDF version of the issue for its exact content.

Findings

Table or figure – see the PDF version of the issue for its exact content.

Explanatory variables examined

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

The gold price exerts the greatest influence on bitcoin returns. This is followed by the oil price. The EPU index in China, the European Union, and the United States negatively affects bitcoin returns.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

A negative relationship was exhibited between liquidity and cryptocurrency returns.

The relationship between Ethereum and Bitcoin is complex and evolving.

EPU has a positive effect on Bitcoin returns.

Investors who account for changes in oil prices are more likely to achieve profits.

The impact of Bitcoin on Ethereum is significantly stronger than the impact of Ethereum on Bitcoin.

Table or figure – see the PDF version of the issue for its exact content.

The daily return of the MSCI World is the only significant variable influencing daily bitcoin returns.

Table or figure – see the PDF version of the issue for its exact content.

The factors influencing the returns of cryptocurrencies across various market...

al. 2019; Zhang–Li 2023), the impact of economic and political instability (Shaikh 2020), oil price fluctuations (Salisu et al. 2023), regulatory news (Auer–Claessens 2018), as well as GDP and unemployment (Corbet et al. 2018) on cryptocurrencies.

Table 1 summarises the most important factors examined in scientific research, as well as the research findings. Based on this, it can be observed that stock market returns, such as the S&P 500 and the MSCI World Index, play a prominent role in the formation of the bitcoin exchange rate. Numerous studies have confirmed that these market indices are closely correlated with changes in bitcoin returns. S&P 500 returns have exerted a significant effect on bitcoin returns in both the short and long term (Ahmed et al. 2023), particularly during the Covid-19 pandemic (Nguyen 2022). In studying the impact of capital market indices, Malladi et al. (2021) also reached the conclusion that the daily return of the MSCI World Index significantly influences the formation of bitcoin returns.

The price of gold is also a frequent variable in the analysis of cryptocurrency returns; some researchers (e.g. Smales 2019) regard bitcoin as a safe haven asset, similar to gold. Panagiotidis et al. (2018) demonstrated a positive correlation between the increase in bitcoin returns and the rise in the gold price. According to Jareño et al. (2020), this positive relationship is primarily observable in the long term. These results support the notion that traditional store-of-value assets, such as gold, can have a significant impact on the cryptocurrency market.

Based on the literature review, it can be established that the VIX index and illiquidity measures, such as the Amihud illiquidity ratio, are also important factors when analysing the exchange rates of bitcoin and other cryptocurrencies. Bouri et al. (2019), for instance, pointed out that the relationship between the VIX index and bitcoin returns is significant, especially during periods of market volatility. In parallel, Liu and Tsyvinski (2018), as well as Eross et al. (2019), demonstrated that illiquidity measures, such as the Amihud illiquidity ratio, play a significant role in the formation of returns for bitcoin and other cryptocurrencies.

The volume of Google searches is another frequently examined influencing factor.

A close correlation can be observed between the volume of Google searches and the bitcoin exchange rate, suggesting that search interest influences prices (Kristoufek 2013). This relationship between Google search trends and bitcoin returns can be observed in both the short and long term, in contrast to the effect of the VIX volatility index, which was only demonstrated in the long term (Chang et al. 2021).

Other factors, such as macroeconomic variables like GDP or the unemployment rate, have also been examined in the literature, though these studies are less frequent. Corbet et al. (2018) demonstrated that news related to unemployment exerts a significant effect on bitcoin returns, whereas news related to GDP and the consumer price index does not.

Another part of the literature examines the impact of cryptocurrencies on one another, pointing out that interactions between different cryptocurrencies can also have a significant effect on exchange rates. For example, Akin et al. (2023) demonstrated that the returns of Ethereum and Binance exert a significant positive effect on the returns of Bitcoin. According to other studies, the relationship between Ethereum and Bitcoin is more complex than this (Sifat et al. 2019; Masiak et al. 2018). According to Masiak et al. (2018), Bitcoin has a stronger effect on Ethereum than Ethereum has on Bitcoin.

The most frequently examined factors influencing cryptocurrency returns in the literature are capital market returns, the returns and prices of precious metals (primarily gold), market risks (such as volatility or illiquidity), as well as news and search frequencies related to cryptocurrencies. Furthermore, it can be observed that previous studies have not examined the impact of these factors across different periods. The aim of this research is to address this gap, namely to contribute to a better understanding of cryptocurrency market dynamics and the formulation of more effective investment strategies through a detailed analysis of different periods.

Based on the literature, the hypotheses are formulated as follows:

H1: There is a significant positive relationship between changes in capital market returns and changes in the returns of the investigated cryptocurrencies.

H2: There is a significant positive relationship between changes in gold returns and changes in the returns of the investigated cryptocurrencies.

H3: There is a significant negative relationship between changes in market risks (the volatility index and illiquidity) and changes in the returns of the investigated cryptocurrencies.

H4: There is a significant positive relationship between the search frequency for Bitcoin and changes in the returns of the investigated cryptocurrencies.

H5: The explanatory power of the investigated factors varies across different periods, with different factors being dominant in each period.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

I divided the full period into three sub-periods in accordance with the structural breaks identified by the Bai–Perron test applied to the bitcoin exchange rate (see Figure 1): 1. between 1 January 2018 and 28 December 2020, 2. between 28 December 2020 and 6 May 2022, and 3. between 6 May 2022 and 31 December 2023.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

where Y is the dependent variable, $\beta_0$ is the intercept of the model, $X_j$ is the $j$-th explanatory variable of the model, $\beta_j$ is the $j$-th regression coefficient ($j=1, p$), and $\varepsilon$ is the error term.

Results

As a first step, I estimated separate regressions for bitcoin returns for the three analysed periods. The explanatory variables were the capital market index (NASDAQ), gold returns, bitcoin search volumes, the volatility index (VIX), and the Amihud illiquidity ratio. Following bitcoin, ethereum and binance are the two largest and most traded cryptocurrencies on the market (Coinmarketcap 2023). In order to provide a more complete picture, I examined whether the effects experienced in the case of bitcoin are also valid for ethereum and binance. This also provided the opportunity to compare these effects across different cryptocurrencies. Table 3: Results of diagnostic tests Tests

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Based on the results of the model diagnostic tests, the results of the re-estimated regression models show (Table 4) that in the first period, gold returns and the volatility index significantly influence bitcoin returns; in the second period, gold returns, the number of searches for bitcoin, the volatility index, and capital market returns are the significant factors; while in the third period, it is the capital market returns that significantly influence bitcoin returns. Similar results were obtained for ethereum and binance.

As shown in Table 4, the coefficient of determination (adjusted R2) of the regressions is relatively low, suggesting that an important factor is missing from the models. According to Katsiampa (2022), bitcoin is the driving force of the cryptocurrency market and influences the returns of other cryptocurrencies. Therefore, I included bitcoin returns among the explanatory variables for ethereum and binance. The results show that in the case of ethereum, bitcoin and capital market returns were significant in the first period (see Table 5); in the second period, bitcoin returns, bitcoin search volumes, and the Amihud illiquidity ratio were significant; and in the third period, bitcoin returns were significant. Similar results were obtained for binance, with the sole exception being the first period, when bitcoin, the VIX, and the Amihud illiquidity ratio were significant influencing factors.

As a summary of the research findings, it can be concluded that the return of the capital market index has a significant positive effect on the returns of the examined cryptocurrencies only in the second and third periods. In the first period, the effect is not significant; thus, the first hypothesis (H1) was only partially supported. Similar results were obtained regarding gold returns: in the first period, a significant positive effect was observed for all three examined cryptocurrencies, but in the second and third periods, there was no significant effect. This suggests that, similar to capital market index returns, the influence of gold is time-varying. Consequently, the second hypothesis (H2) was also only partially supported.

Market risks, the VIX, and the Amihud illiquidity ratio exhibit a negative significant effect on the returns of the examined cryptocurrencies. In this case as well, the effect varies across the different periods. Therefore, the third hypothesis (H3) was also only partially supported. The number of searches for bitcoin showed a significant effect only in the second period, meaning the fourth hypothesis (H4) was also only partially supported. It can be observed that this effect is considerably stronger for ethereum and binance than for bitcoin; however, since the coefficient value is very low in all cases, it can be said that the effect is marginal even in the second period.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

Table or figure – see the PDF version of the issue for its exact content.

The preceding findings demonstrate that the exchange rates of the examined cryptocurrencies were influenced by different factors across various periods, and that the strength of the impact of each factor varies by period. In the first period, the gold price and volatility predominated, whereas in the second period, the stock market index and search volumes became decisive. In the third period, the stock market index was the significant factor. Based on these findings, it can be concluded that the fifth hypothesis (H5) has been proven, i.e., it is worthwhile to conduct the analysis across different periods.

Conclusions

Based on the results of the research, the most important conclusion to be drawn is that the factors influencing the dynamics of the cryptocurrency market change over time and appear in different orders of importance across various periods; therefore, it is advisable in future research to examine each period individually.

Furthermore, it can be observed that during the first period, which was characterised by relative stability and calm in the cryptocurrency market, the returns of Bitcoin, Ethereum, and Binance were most influenced by gold returns and the volatility index. This is in accordance with the findings of Panagiotidis et al. (2018). When the market struggles with turbulence (see the second period), the number of searches for Bitcoin and stock market returns became the most important influencing factors. These results are also supported by the study of Lahmiri et al. (2020). This period is a time of market tension and uncertainty, during which investors' attention is primarily directed towards information related to Bitcoin and changes in stock market returns. In the third period, when the market shows a decline, stock market returns became the most dominant influencing factor. It can be observed that changes in the returns of Ethereum and Binance are largely explained by changes in Bitcoin returns. This is consistent with the findings of Masiak et al. (2018) and Katsiampa (2022).

As a limitation of the research, it is important to emphasise that only three cryptocurrencies were examined during the analysis; thus, the results are not necessarily representative of the cryptocurrency market. Another limitation of the research is that there was only one bubble period during the examined timeframe (2018–2023), so it would be worthwhile to repeat the research after the next bubble period. In this study, I only employed the OLS (ordinary least squares) regression method. Although this is a widespread and reliable method, more detailed and accurate results could be obtained by applying more complex models.

Bibliography

Ahmed, M. Y.–Sarkodie, S. A.–Leirvik, T. 2023. Mutual coupling between stock market and cryptocurrencies. Heliyon 9(5), e16179. https://doi.org/10.1016/j.heliyon.2023. e16179 Akin, I.–Khan, M. Z.–Hameed, A.–Chebbi, K.–Satiroglu, H. 2023. The ripple effects of CBDC-related news on Bitcoin returns: Insights from the DCC-GARCH model.

International Business and Finance 66, 102060. https://doi.org/10.1016/j.ribaf.2023.102060 Auer, R.–Claessens, S. 2018. Regulating cryptocurrencies: assessing market reactions.

BIS Quarterly Review. https://ssrn.com/abstract=3288097 Baker, S. R.–Bloom, N.–Davis, S. J. 2016. Measuring economic policy uncertainty. The Quarterly Journal of Economics 131(4), 1593–1636. https://doi.org/10.1093/qje/qjw024 Bloomberg 2021. Bitcoin went mainstream in 2021. It’s just as Volatile as ever. https://www.bloomberg.com/graphics/2021-bitcoin-volitility/, accessed: 23. 10. 2023.

Burton, A. L. 2021. OLS (Linear) regression. The encyclopedia of research methods in criminology and Criminal Justice 2, 509–514. https://doi.org/10.1002/9781119111931. ch104 Chang, C. Y.–Lo, C. C.–Cheng, J. C.–Chen, T. L.–Chi, L. Y.–Chen, C. C. 2021.

Relationship between the Bitcoin exchange rate and other financial indices in time series.

Mathematical Problems in Engineering 1, 8842877. https://doi.org/10.1155/2021/8842877 Coinmarketcap 2023. Today’s cryptocurrency prices by market cap. https://coinmarketcap.com, accessed: 28 Dec 2023.

Corbet, S.–Larkin, C. J.–Lucey, B. M.–Meegan, A.–Yarovaya, L. 2018. The volatility generating effects of macroeconomic news on cryptocurrency returns. https://ssrn.com/abstract=3141986 Das, D.–Kannadhasan, M. 2018. Do global factors impact bitcoin prices? Evidence from wavelet approach. Journal of Economic Research 23(3), 227–264.

Eross, A.–McGroarty, F.–Urquhart, A.–Wolfe, S. 2019. The intraday dynamics of bitcoin. International Business and Finance 49, 71–81. https://doi.org/10.1016/j.ribaf.2019.01.008 Huynh, T. L. D.–Burggraf, T.–Wang, M. 2020. Gold, platinum, and expected Bitcoin returns. Journal of Multinational Financial Management 56, 100628. https://doi.org/10.1016/j.mulfin.2020.100628 Jareño, F.–de la O González, M.–Tolentino, M.–Sierra, K. 2020. Bitcoin and gold price returns: A quantile regression and NARDL analysis. Resources Policy 67, 101666. https://doi.org/10.1016/j.resourpol.2020.101666 Katsiampa, P.–Yarovaya, L.–Zięba, D. 2022. High-frequency connectedness between Bitcoin and other top-traded crypto assets during the COVID-19 crisis.

Journal of International Financial Markets, Institutions and Money 79, 101578. https://doi.org/10.1016/j.intfin.2022.101578 Kristoufek, L. 2013. BitCoin meets Google Trends and Wikipedia: Quantifying the relationship between phenomena of the Internet era. Scientific Reports 3(1), 3415. https://doi.org/10.1038/srep03415 Lahmiri, S.–Bekiros, S.–Bezzina, F. 2020. Multi-fluctuation nonlinear patterns of European financial markets based on adaptive filtering with application to family business, green, Islamic, common stocks, and comparison with Bitcoin, NASDAQ, and VIX. Physica A: Statistical Mechanics and its Applications 538, 122858. https://doi.org/10.1016/j.physa.2019.122858 Malladi, R. K.–Dheeriya, P. L. 2021. Time series analysis of cryptocurrency returns and volatilities. Journal of Economics and Finance, 45(1), 75–94. https://doi.org/10.1007/s12197-020-09526-4 Masiak, C.–Block, J. H.–Masiak, T.–Neuenkirch, M.–Pielen, K. N. 2018. The triangle of ICOs, Bitcoin and Ethereum: A time series analysis. https://hdl.handle.net/10419/207030

Nguyen, K. Q. 2022. The correlation between the stock market and Bitcoin during COVID-19 and other uncertainty periods. Finance Research Letters 46, 102284. https://doi.org/10.1016/j.frl.2021.102284 Panagiotidis, T.–Stengos, T.–Vravosinos, O. 2018. On the determinants of bitcoin returns: A LASSO approach. Finance Research Letters 27, 235–240. https://doi.org/10.1016/j.frl.2018.03.016 Salisu, A. A.–Ndako, U. B.–Vo, X. V. 2023. Oil price and the Bitcoin market. Resources Policy 82, 103437. https://doi.org/10.1016/j.resourpol.2023.103437 Shaikh, I. 2020. Policy uncertainty and Bitcoin returns. Borsa Istanbul Review 20(3), 257–268. https://doi.org/10.1016/j.bir.2020.02.003 Sifat, I. M.–Mohamad, A.–Shariff, M. S. B. M. 2019. Lead-lag relationship between bitcoin and ethereum: Evidence from hourly and daily data. International Business and Finance 50, 306–321. https://doi.org/10.1016/j.ribaf.2019.06.012 Smales, L. A. 2019. Bitcoin as a safe haven: Is it even worth considering?. Finance Research Letters 30, 385–393. https://doi.org/10.1016/j.frl.2018.11.002 Zhang, W.–Li, Y. 2023. Liquidity risk and expected cryptocurrency returns. International Journal of Finance & Economics 28(1), 472–492. https://doi.org/10.1002/ijfe.2431 Ødegaard, B. A. 2018. Liquidity–The Amihud Illiquidity Estimator, Academic Press.

Authors

  • Kovács Edina

Text extracted automatically from the PDF; for tables, figures and exact layout see the PDF of the issue.