Effects of COVID-19 on cryptocurrency and emerging market connectedness: Empirical evidence from quantile, frequency, and lasso networks

Abstract

We use time and frequency connectedness approaches based on network analysis to investigate the volatility connectedness among 27 emerging equity markets and seven high-capitalized cryptocurrencies. We estimate the network connectedness using the standard, quantile, frequency, and lasso VAR models for the pre- and post-COVID-19 pandemic periods and daily data over the period from October 2, 2017 to May 20, 2022. The network connectedness estimates based on the several models used in this study indicate a growing risk spillover among and within the emerging market equities and the cryptocurrencies after the COVID-19 pandemic hit the world.  The frequency connectedness analysis shows that cryptocurrencies cannot be used as diversifiers for emerging stock markets in both the short and long-run. The empirical findings from the quantile VAR model reveal that the volatility connectedness in the tails is much stronger compared to the center of the distribution. It is also evident that Saudi Arabia, Thailand’s stock markets, and USDT are the main risk transmitters at the 0.95-th quantile during the post-COVID period. Time-varying connectedness estimates confirm the substantial effect of COVID-19. Our study also shows that the spread of risk among these financial markets is global rather than regional, supporting cross-border structure and worldwide financial market integration. The findings suggest cryptocurrency and emerging market equity portfolios should be closely monitored during financial turmoil.

Introduction

The COVID-19 (SARS-CoV-2) pandemic, also known as the novel coronavirus pandemic, started around the end of December 2019 in China and quickly accelerated, reaching every corner of the globe in just a few months. According to the World Health Organization (WHO), there have been nearly 274.6 million confirmed cases, and 5.4 million people have died as of December 18, 2021. Although its destructive effect is slowly decreasing as the scope of vaccination expands, governments have continued to impose a range of ‘lockdown’ measures such as a combination of stay-at-home orders, travel bans, closing schools, and nonessential business restrictions on public and private gatherings. The COVID-19 pandemic has caused a catastrophic loss of life, but it has also caused a global economic downturn so far, and still, its effects persist. No previous epidemic, including the 1918 Spanish Flu, has shaken stock markets as strongly as COVID-19. For example, the volatility index of the S&P500 jumped dramatically about 500% from January 15, 2020, to March 31, 2020 [1]. Additionally, the global stock markets have exhibited varying degrees of volatility following the pandemic. This unprecedented shock has caused severe falls in stock markets around the world. In addition to the stock markets, COVID-19 also hit the cryptocurrency markets very hard, and most cryptocurrencies lost half their value. For instance, the price of Bitcoin fell from 9000 US dollars to around 4000 US dollars within days after March 7, 2020. Although cryptocurrencies have several potential benefits, their volatile price dynamics could be detrimental to market participants [2], [3]. During this turmoil, investors have attempted to diversify their portfolios towards​ cryptocurrencies to take advantage of short-term gains [4]. Thus, the novel coronavirus has already exerted significant consequences on the global financial sector, encouraging finance scholars to investigate its impacts on financial markets [5]. For example, Caferra and Vidal-Tomás [6], James [7], Corbet et al. [8], Conlon et al. [9], and Lahmiri and Bekiros [10] have recently examined the effects of COVID-19 on various financial markets and portfolio selection. A few other studies consider the highly volatile and frequently occurring abrupt changes in cryptocurrency markets and use methods from the econophysics literature to analyze the statistical properties of cryptocurrency prices and their comovement with the prices of other assets using high frequency data [7], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22].

In recent years, crypto assets have become increasingly popular among investors. Since its introduction in 2009, the overall market capitalization of crypto assets has expanded exponentially, from less than 20 billion US dollars in January 2017 to more than 3 trillion US dollars by the end of 2021. The debate over the role of cryptocurrencies in the economy is still unresolved. Despite the fact that cryptocurrencies are used for payments, their function as a medium of exchange is questioned [23], [24]. Some authors, on the other hand, argue that cryptocurrencies, like gold, silver, and other financialized metals, can be considered physical goods or commodities. Cryptocurrencies, according to Dyhrberg [25], have characteristics that are very similar to gold and the dollar. Investors may include cryptocurrencies in their portfolios for diversification, hedge, or safe-haven purposes due to these characteristics of cryptocurrencies as financial assets. Because of their low correlation with stock markets, cryptocurrencies can act as diversifiers [26], reducing volatility risks by smoothing out unsystematic risk. Cryptocurrencies have been found to have no significant correlation or a negative correlation with some equities [13], suggesting that they could be used as a hedge instrument in some portfolios. Investors take opposing positions in hedge assets to reduce the potential losses due to price fluctuations, especially during severe financial crises. This is especially important for emerging markets, as the instruments available for hedging risks in emerging markets are limited in comparison to advanced markets [27]. A safe-haven instrument is a hedge asset that has a negative correlation with the portfolio’s main assets and is used to offset the risk of the existing portfolio during financial crises.

Considering interconnectedness with other assets and their huge market capitalization, crypto-assets could pose a risk to the financial system; spillovers may also affect the real economy. The use of crypto-assets, in particular, has the potential to adversely affect financial stability, interfering with payments and market infrastructures, and influencing monetary policy. This study focuses on the relationship between the cryptocurrency market and emerging stock markets, as cryptocurrencies are predominantly adopted by emerging markets and their potential to cause financial instability should be more closely linked to emerging markets. Indeed, nineteen of the top twenty countries in the Global Crypto Adoption Index (CAI) published by Chainanalysis in 2021 [28] are emerging economies. However, emerging market adoption of cryptocurrencies is far higher than in advanced economies [29], meaning that emerging markets are more exposed to cryptocurrency risk. This comes as a result of recent bouts of high price volatility, which have aroused concerns about their possible impact on financial stability. A recent report by the International Monetary Fund [30] warns that a surge in the trade of cryptocurrencies in emerging economies could destabilize the global financial system. Wider adoption of cryptocurrencies by emerging markets and greater financial instability risks the interconnectedness of them are related to several crucial elements. First, compared to the advanced markets, emerging markets have limited hedging options and short selling is not allowed in many emerging markets, which limits their hedging [27], [31]. The significance of emerging market risk is greater than it appears, given that advanced economies have substantial portfolio investments in emerging markets and emerging market economies are largely driven by the developments of advanced economies. According to the Institute of International Finance [32], portfolio flows to emerging markets in December 2021 totaled 16,8 billion US dollars. In addition, emerging markets are vulnerable to numerous political, geopolitical, conflict, and natural occurrences. Consequently, hedging emerging market risk is essential for portfolio investors not only in emerging markets but globally as well. Consequently, our study investigates whether emerging market risks can be hedged using cryptocurrencies and how events such as the COVID-19 pandemic impact hedging capacity.

Against this backdrop, the objectives of this paper are as follows: First, the study aims to determine whether events leading to significant volatility jumps cause significant changes in the connectedness of cryptocurrencies among themselves and, more importantly, their connectedness with emerging market stocks. We do this by examining the effect of the COVID-19 pandemic, which caused a large volatility jump in both cryptocurrency and emerging stock markets, on the connectedness of the cryptocurrencies with a large number of emerging market stocks in terms of intraday volatility connectedness. A second objective of the study is to investigate how connected movements in cryptocurrency and emerging stock markets are. We also investigate whether the risk spillover among cryptocurrencies and emerging stock markets has changed over time, particularly how the COVID-19 pandemic affected this. A third objective is to convey a deeper understanding of risk spillover structures among the markets considered using a number of network-based metrics, such as the in- and out-degree, centrality, betweenness, and page rank. Finally, a fourth objective is to investigate whether risk connectedness among emerging stock markets and cryptocurrencies changes in response to the effects of the COVID-19 pandemic, and thus, to identify whether cryptocurrencies can be robustly used as portfolio diversifying instruments for emerging stock market risks.

Given its importance and suitability for measuring spillovers, we use the volatility network connectedness approach to analyze the risk transmission among 27 emerging equity markets and seven leading cryptocurrencies with high market capitalization. We demonstrate risk transmission patterns among these financial markets before and during the COVID-19 pandemic using several network connectedness measures. For the analysis, we use daily frequency data covering the period from October 2, 2017 to May 20, 2022 with 981 observations. The contribution of this study to the existing literature is threefold. First, we extend the cryptocurrency literature by incorporating all emerging equity markets during and after the COVID-19 pandemic. In so doing, we examine how COVID-19 shifts the risk transmission channels of such risky markets at the global level. Second, we evaluate the diversification benefits of high-capitalized seven major cryptocurrencies for emerging stock markets by considering various market conditions. In addition to mean-based risk spillovers, we also consider short- and long-term risk distributions as well as tail risk distributions. Third, we calculate several network measures, such as the centrality, betweenness, and degree, to detect financial markets that play a critical role in risk transmission during the pre- and post-COVID-19 outbreak periods. We also use alternative robust estimation approaches such as the quantile VAR, frequency VAR, and Big VAR Lasso for robustness. Hence, our results provide valuable information for stock market investors who want to diversify their risky assets with cryptocurrencies.

We examine the volatility spillover connectedness among the related equity markets and cryptocurrencies by combining the forecast error variance decomposition (FEVD) framework of Diebold and Yılmaz [33] and the network theory. In addition to a standard vector autoregressive (VAR) model, we also employ the frequency connectedness approach of Baruník and Křehlík [34] to assess the effect of risk spillovers on the related financial markets at different frequency bands, i.e., short and long-term investment horizons. To move beyond the mean-based VAR framework and capture connectedness under extreme events, we employ a quantile-based connectedness analysis using estimates from a quantile VAR model, which is based on the quantile regression of Koenker and Bassett Jr. [35]. Finally, we also employ a big VAR model, with shrinkage and selection via the lasso approach of Tibshirani [36], to check the robustness for high dimensionality issues as our analysis uses a relatively large number of variables. Third, we examine several network-based centrality measures such as in-degree centrality, out-degree centrality, eigenvector centrality, betweenness, and PageRank centrality to gain insights into the risk transmission mechanism before and after the outbreak of COVID-19.

The empirical results from the different approaches suggest a growing risk transmission both within and across emerging stock markets and cryptocurrencies after the outbreak of the COVID-19 pandemic. This is robust evidence to support that COVID-19 increased the volatility spillover and changed the direction of risk. Network centrality scores show that cryptocurrencies cannot be used as diversifiers for emerging stock markets due to their higher scores during both pre- and post-COVID periods. Given that financial markets’ extreme fluctuations show similar dynamics during the COVID-19, the model based on the tail risk measurements may provide more accurate outcomes for the post-COVID-19 period. Thus, the empirical findings show that financial system volatility flows towards certain assets prior to and during the post-COVID period. For instance, the stock markets of Saudi Arabia, Thailand, and USDT are the leading risk transmitter network nodes at 0.95 quantiles during the post-COVID period. Lastly, the cluster analysis obtained from various models also shows that the number of clusters substantially decreases after the COVID-19 pandemic. This can be shown as evidence that the pandemic put the financial markets in the same risk cluster and brought them closer to each other.

The remainder of the paper has the following structure: In Section 2, we provide a review of the related literature. Section 3 describes the various models and the measures of connectedness used in the paper. Section 4 presents the data and some descriptive statistics. Section 5 reports our empirical findings and offers a discussion of the results, and the last section concludes our paper.