Is cryptocurrency a hedging tool during economic policy uncertainty? An empirical investigation

Recently, the fascination with cryptocurrencies has been growing, and Bitcoin has emerged as the most notable. Cryptocurrency, in its broadest sense, refers to a form of digital currency that operates on the Internet and independently of any central governing body. Whether cryptocurrencies can serve as a medium of exchange has been the subject of some research, and the results have been mixed. Due to the dramatic surge in cryptocurrency values in 2017, investors worldwide began pouring much of their investment capital into these relatively new forms of financial assets. However, the inflated prices could not be maintained, and the cryptocurrency market experienced a price bubble bust before the year's conclusion. The enormous volatility in the value of cryptocurrencies demonstrates the risk involved in investing in this kind of asset. Cryptocurrencies are established through various cryptographic algorithms and are exchanged in a digital realm. The ongoing development of this system, which aims to replace existing currencies, payment instruments, and even traditional monetary theory and practices, has increased the significance of this system over time. In contrast, there is a debate among scholars and experts about whether cryptocurrency is a type of money or a volatile asset. Opinions differ on this phenomenon. Global financial markets have seen a rapid increase in cryptocurrency popularity. As a result, regulators, the media, and individual and institutional investors have all taken an interest in them. Academic cryptocurrency research has also become significant. A feature of both gold and US dollars, Bitcoin's potential to hedge is known as a medium of exchange or digital gold. Because a positive link exists between the prices of gold and Bitcoin, gold and Bitcoin are more likely to complement than compete against one another. A short position in the Bitcoin market enables hedging the risk when investing in various other financial assets. In particular, the portfolio's risk from holding gold, oil, and stocks is lower when holding Bitcoin than when not. However, because of restrictions on anti-money laundering and terrorist funding legislation, Bitcoin cannot replace gold. Cryptocurrency markets can be affected in various predictable ways. Existing studies have examined how different uncertainty metrics affect cryptocurrency. Some prevailing studies on Bitcoin have examined the impact of uncertainties and risks on cryptocurrencies' profits and price volatility. Prior findings show cryptocurrency market volatility tends to increase in response to heightened investor apprehension; some uncertainty measures adversely affect long-term fluctuations of cryptocurrencies; and trade policy uncertainty can negatively impact Bitcoin returns. Because the modern, financially connected world is more vulnerable to economic policy risk than ever, researchers are currently concentrating on finding an appropriate shelter to protect assets. Previous studies have been conducted on cryptocurrency, mainly on Bitcoin as a single entity with different uncertainty-related measures. To our knowledge, no earlier studies on cryptocurrency with economic policy uncertainty or on the asymmetric effect of EPU on cryptocurrency returns have been done. The study's goals are twofold: to better understand the academic literature already available on crypto investor behavior, compile its knowledge, and identify knowledge gaps to support future studies; and to present significant research findings for investors, academics, policy-makers, businesses, professionals, and society. To our knowledge, this study is the first to analyze the nexus between cryptocurrency and economic policy uncertainty after Covid-19, focusing on the top three cryptocurrencies (Bitcoin, Ethereum, and Tether) from January 2021 to April 2023. Significance of the study. Global financial markets are in a revolutionary phase, and digital finance plays a significant role in how financial services are organized worldwide. Cryptocurrency is considerably improving and moderating traditional financial services, but the market is marked by anomalous behavior and unanticipated occurrences that influence people's views, market behavior, and public legislation. Uncertainty significantly impacts the transmission of fiscal and monetary policies in financial markets. Regulating cryptocurrency is necessary since it alters typical financial transactions; however, keeping up with the legislation in many jurisdictions is challenging. Since their inception, cryptocurrencies have been popular in the financial industry, and the associated markets have a history of volatility. We use a quantile regression approach to investigate cryptocurrencies with global economic policy uncertainty, using Bitcoin, Ethereum, and Tether, and check robustness with gold as an alternative hedge. We also conduct country-specific analysis for the US and China, two key economies with high cryptocurrency adoption. Growing global EPU has a detrimental effect on Bitcoin's long-term returns, while declining EPU has a favorable impact, suggesting investors recover trust as concerns ease. Tether benefits from rising EPU due to its stablecoin status, indicating potential haven properties in the long term, while Bitcoin and Ethereum show short-term haven potential. Motivation of the study. Pandemic containment measures in 2020 shocked economic activity and liquidity, motivating investors and fund managers to adopt risk-reduction strategies. Cryptocurrency, as a pioneering technology, may alter conventional financial institutions. There remains disagreement about whether cryptocurrencies fulfill the three functions of money. Prior literature shows mixed evidence on predictability of Bitcoin returns. Therefore, this study analyzes the nexus between economic policy uncertainty and cryptocurrency returns.

After the introduction of Bitcoin in 2009, cryptocurrencies became accessible to institutional and individual investors and have been increasingly studied. Contemporary research spans technological underpinnings and economic/financial ramifications, with a persistent inquiry: do cryptocurrencies function as a medium of exchange or as financial assets? Many studies examine whether Bitcoin or other cryptocurrencies can serve as a hedge similar to gold during uncertain times, but findings diverge due to differences in periods, drivers, methods, and assets considered. Early work (e.g., Bouri et al., 2017) found Bitcoin hedging more prevalent in both bullish and bearish market conditions at shorter horizons under global uncertainty, with OLS showing a negative correlation. Demir et al. (2018) used EPU as uncertainty metric and found EPU could forecast Bitcoin returns, with a negative relation overall but positive effects in upper quantiles, indicating hedging in bull markets. When accounting for EPU, Bitcoin’s hedging utility can improve for bond and equity portfolios under specific conditions. Some works challenge safe-haven characteristics (e.g., Mokni, 2021; Wu et al., 2019), while others support them. Theoretical discourse addresses whether cryptocurrencies can be money given lack of central management, fluctuating demand, and fixed supply. Adoption has grown, with many corporations accepting Bitcoin, yet debates continue about its monetary attributes versus asset characteristics. EPU, developed by Baker et al. (2016), is a widely used proxy for economic uncertainty, with both direct channels (policy decisions impacting crypto markets) and indirect channels (erosion of confidence in fiat economies) influencing cryptocurrency demand. For example, regulatory announcements have triggered immediate Bitcoin price declines. Elevated EPU can drive investors to alternative assets, positioning cryptocurrencies as potential hedges during uncertainty.

Data and sample: Monthly data from 1 January 2021 to 1 April 2023 were used to focus on the post-COVID-19 period. Cryptocurrency returns (BTC, ETH, THT) were sourced from coindesk.com; economic policy uncertainty (global and country-specific EPU indexes) from policyuncertainty.com; and gold returns from investing.com. Interest is included as a control variable. Model: CR_t = β0 + β1 EPU_t + GRT_t + CV_t + μ_t, where CR denotes cryptocurrency return at time t (BTC, ETH, or THT), EPU_t denotes the EPU index at time t, GRT_t represents gold returns, CV are control variables (including interest), and μ_t is the error term. Variables: Cryptocurrency returns for Bitcoin, Ethereum, and Tether proxy crypto performance. EPU captures uncertainty around future government policy and regulation that may induce delays in spending and investment. Gold returns serve as a benchmark haven asset. Estimation procedures: The study primarily employs Quantile Regression (QR; Koenker and Bassett, 1978) to estimate conditional quantiles of cryptocurrency returns as functions of EPU and controls, which is robust to non-normal errors, heteroskedasticity, and outliers, and captures heterogeneous effects across the return distribution. The formal definition of quantiles and the check-loss minimization framework are outlined. Preliminary data treatment includes unit root/stationarity checks via ADF tests (Table 1), summary statistics (Table 2), and pairwise correlations (Table 3) to assess multicollinearity risk. Granger causality tests (Table 4) assess predictive relations between EPU and assets. Robustness and extensions: Gold is used as an alternative hedge comparator. A composite cryptocurrency index (CR) is constructed via principal component analysis (PCA) combining BTC, ETH, and THT, and subjected to QR (Table 6). Country-specific QR analyses assess the impact of US and China EPU on Bitcoin returns (Table 7). Asymmetric dynamics: To capture potential short- and long-run asymmetries, the Nonlinear Autoregressive Distributed Lag (NARDL) framework is applied. The EPU series is decomposed into positive (EPU+) and negative (EPU−) partial sum processes, allowing separate estimation of effects from increases and decreases in EPU on crypto returns. The approach is well-suited for small samples, yields short-run and long-run coefficients, and mitigates residual correlation concerns. Long-run coefficients (Table 8) and short-run dynamics (Table 9) are estimated, and Wald tests assess short-run asymmetry. Heterogeneity across cryptocurrencies is examined via slope equality and symmetry quantile tests (Table 10).

- Granger causality (Table 4): Bidirectional predictability between EPU and ETH (ETH → EPU: F=5.8610, p=0.0116; EPU → ETH: F=4.1159, p=0.0305). For BTC, evidence is weaker but suggestive (BTC → EPU: F=3.3586, p=0.0590; EPU → BTC: F=2.3697, p=0.0964). No significant Granger causality between EPU and Gold. - Global quantile regression (Table 5): Effects of EPU on BTC and ETH are heterogeneous across quantiles. For BTC, significant positives at q=0.40 (coef=0.00569, p<0.05) and q=0.80 (coef=0.00252, p<0.05), but significant negatives at q=0.50 (coef=-0.00578, p<0.10) and q=0.90 (coef=-0.00234, p<0.10). For ETH, significant positive at q=0.40 (coef=0.00432, p<0.05) and at q=0.80 (coef≈0.02112, p<0.10), and significant negative at q=0.90 (coef=-0.00325, p<0.05). For THT, most coefficients are small and insignificant across quantiles, indicating limited sensitivity to EPU. - Robustness with gold and composite CR (Table 6): Gold shows a significant positive relation with EPU at q=0.90 (coef=0.109, p<0.05). The composite crypto index (CR via PCA) shows significant positives at q=0.50 (coef=0.00912, p<0.05) and q=0.60 (coef=0.00937, p<0.05), suggesting mid-quantile sensitivity. - Country-specific QR (Table 7): China EPU coefficients are mostly negative and statistically insignificant across quantiles, with significant adverse effects at higher quantiles (q=0.80: coef=-0.627, p<0.10; q=0.90: coef=-0.748, p<0.01), suggesting limited or adverse impact on BTC returns. US EPU exhibits significant positive effects at mid quantiles (q=0.30: 0.852, p<0.01; q=0.40: 1.166, p<0.05; q=0.50: 1.211, p<0.05) and significant negatives at q=0.60 (−1.414, p<0.05) and q=0.70 (−0.994, p<0.10), with a positive at q=0.80 (0.859, p<0.10), indicating nuanced, quantile-dependent impacts consistent with the US’s global policy influence. - NARDL long-run (Table 8): BTC exhibits significant negative long-run elasticities to both EPU increases (EPU+ coef=−5.3392, p<0.01) and decreases (EPU− coef=−4.6494, p<0.01), implying rising EPU depresses BTC long-run returns, while interpretation within the asymmetric framework indicates returns rise when uncertainty declines over time. ETH long-run coefficients are reported as negative (EPU+) and positive (EPU−), though without statistical markers in the table snippet. THT shows significant positive long-run associations with both EPU+ (coef=0.8606, p<0.05) and EPU− (coef=1.886, p<0.05), indicating long-term alignment with EPU and potential haven-like behavior for a stablecoin. - NARDL short-run (Table 9): Positive EPU shocks raise BTC and ETH in the short run at various lags (e.g., BTC ΔEPU+ contemporaneous 7.702, p<0.05; lag-1 10.650, p<0.10; lag-2 24.79, p<0.05; lag-3 16.666, p<0.01). Negative EPU shocks (declines) often have larger magnitude effects, with significant negative coefficients implying increases in returns when uncertainty declines. THT shows mixed short-run responses, including negative returns at various lags under rising uncertainty, suggesting it is not the immediate short-term hedge of choice. Wald tests support short-run asymmetry. - Overall: EPU has little effect on cryptocurrencies in the short term on average, but quantile- and asymmetry-aware analyses reveal that BTC and ETH can act as short-term hedges during elevated uncertainty, whereas their long-run safe-haven properties diminish. Tether displays a positive long-run association with EPU, consistent with its stablecoin nature.

The study asks whether cryptocurrencies hedge economic policy uncertainty. The quantile regression results demonstrate heterogeneous and state-dependent relationships: at higher or specific quantiles, Bitcoin and Ethereum exhibit hedging characteristics against EPU, supporting their role as short-term hedges during periods of elevated uncertainty. However, the NARDL long-run estimates indicate that increases in EPU reduce the long-run returns of BTC and ETH, suggesting these assets are not reliable long-term safe havens as policy uncertainty rises. Conversely, Tether’s positive long-run association with EPU aligns with its stablecoin design and safe-haven-like behavior. Country-specific analyses highlight the outsized role of US policy uncertainty in shaping global cryptocurrency dynamics: US EPU significantly affects BTC returns across several quantiles, while China’s EPU shows largely insignificant or adverse effects, underscoring differing transmission channels and market integration. These findings address the research question by indicating that cryptocurrencies—particularly BTC and ETH—can provide short-term hedging benefits during high EPU, but their effectiveness wanes in the long run. The results are relevant for portfolio diversification, risk management, and policy, as heightened EPU amplifies information asymmetries and investor heterogeneity, and the US policy environment has disproportionate spillovers onto crypto markets.

Using monthly post-COVID-19 data (Jan 2021–Apr 2023) and quantile regression, Granger causality, and NARDL, the study finds that BTC and ETH can hedge short-term economic policy uncertainty at specific return quantiles, but their long-run returns are negatively affected by rising EPU, indicating limited safe-haven reliability over longer horizons. Tether exhibits positive long-run associations with EPU, consistent with safe-haven behavior typical of a stablecoin. Robustness checks with gold and a composite crypto index reinforce these insights. Policy-makers should note that US policy uncertainty exerts significant influence on cryptocurrency markets, and the crypto market’s susceptibility to policy-induced risk necessitates careful consideration of regulatory actions. The study contributes to the literature on crypto as a hedge under policy uncertainty and encourages further investigation into market mechanisms and cross-country dynamics. Policy implications: Rapid declines in crypto asset values can adversely affect investors and expose traditional financial institutions through direct and indirect linkages. Stablecoin runs, if poorly regulated, can trigger reserve liquidations and spillovers. Widespread adoption of crypto assets, particularly stablecoins, may disrupt bank deposits and credit intermediation in vulnerable financial systems. Future research directions: Extend country coverage to assess whether short-term hedging properties generalize beyond the US; examine interactions between emerging central bank digital currencies and cryptocurrencies; and expand economic impact assessments of cryptocurrencies beyond environmental concerns.

The country-specific interpretations are presented with caution; the authors note that these are hypotheses and that other variables may influence the observed patterns. The early-stage nature of cryptocurrency markets and varying national regulatory responses also suggest that results may be context-dependent and warrant further investigation.