Since the inception of Bitcoin in 2008, crypto assets, have long been perceived as one of the riskiest assets.1 Unlike traditional assets, many cryptocurrencies are decentralized, limited in supply, and difficult to value, leading to speculation, high divergence of opinion, and extreme price volatility. These features, combined with the rapid growth of crypto markets during the COVID-19 pandemic and the increasing linkage of crypto assets to the financial system, understandably give rise to regulatory concerns. One such concern is that with the rising comovement between crypto markets and stock markets, the wild volatility in the crypto markets may pose contagion risk to other markets and trigger stress in the entire financial system. For example, a recent report by the International Monetary Fund (IMF) warns that crypto assets “could pose financial stability risks given [their] highly volatile prices, the rising use of leverage in their trading, and financial institutions’ direct and indirect exposures to these assets.” The regulatory concerns about crypto assets have been heightened in the turmoil following the crash of algorithmic stablecoin TerraUSD, which wiped out over $50 billion, and the collapse of FTX, the second-largest crypto exchange at the time, which shook crypto investors’ confidence more broadly.
Despite the risky nature of crypto assets, there are reasons to believe that volatility in crypto markets does not necessarily spill over to stock markets. First, while regulators repeatedly stress that cryptocurrencies are no longer obscure assets on the fringe of the financial system, the fact remains that crypto markets are a fraction of stock markets. Second, crypto traders are no strangers to bear markets and adverse shocks. There have been five “crypto winters” since 2017 and three since 2021, but none of the past crypto winters triggered systemic risk. Furthermore, Bitcoin price volatility has declined in recent years (Liu and Tsyvinski 2021), and may continue to do so as crypto markets grow and attract more institutional investors.
In our recent study, we conduct a close investigation of the evolving correlation between crypto and stock markets, with particular emphasis on the factors driving the two markets’ comovement, the potential for risk spillovers between the two, and the role of institutional involvement.
We present three important findings. First, we find that the return correlation between Bitcoin and the S&P 500 index, the two most followed barometers of the crypto and stock markets, respectively, hovered around zero before March 2020 but exhibited a significant structural increase after. This pattern extends to other market and industry equity indices. The jump in the Bitcoin-S&P 500 return correlation is sudden rather than gradual. The timing of the shift is also inconsistent with an isolated shock to crypto markets generating risk spillovers. Rather, the shift coincides with the unprecedented interventions made by the Federal Reserve (the “Fed”) in response to the market turmoil brought about by the coronavirus outbreak. Indeed, we show that spikes in the crypto-stock correlation align well with dates of the Federal Open Market Committee (“FOMC”) meetings and interest rate changes during the pandemic. The largest jump in the BTC-S&P 500 return correlation appears to be induced by the two unscheduled FOMC meetings held in March 2020, during which the federal funds rate was cut to near zero. A series of FOMC meetings starting in March 2022 also helped sustain a high crypto-stock correlation when the Fed started raising interest rates. These results suggest that the shift in the crypto-stock correlation is most likely fueled by the Fed’s pandemic response, which affected investors’ risk appetite by altering interest rates.
Second, we find little evidence of shocks to crypto markets being transmitted to stock markets. We also find that the S&P 500 index is barely affected by catastrophic events in the crypto markets. On the contrary, using both granger causality tests and impulse response functions, we find that unit shocks to the volatility of S&P 500 returns generate significant and persistent responses in the volatility of Bitcoin returns. Thus, despite concerns about systemic risk triggered by crypto markets, our evidence is more consistent with risk spillovers, if any, extending from stock markets to crypto markets than the reverse.
Third, we identify the growing presence of institutional involvement in the crypto markets as a mechanism through which monetary policy changes affect the crypto-stock comovement. To capture institutional involvement, we study trends in the number of institutions owning a position in the Grayscale Bitcoin Trust (“GBTC”) and trends in the “Coinbase Premium Index.” This index, motivated by the fact that Coinbase serves more institutional clients than Binance, measures the difference in the spot price of Bitcoin between Coinbase and Binance, so a more positive (negative) value suggests higher buying (selling) pressure from institutions. We show that the number of institutions holding Bitcoin, either indirectly through GBTC or directly through Coinbase (as implied by the Coinbase Premium Index), increased sharply after the Fed’s rate cuts during the pandemic but fell after the Fed began rate hikes. We further show that the absolute value of the Coinbase Premium Index, which measures whether buying/selling pressure is more likely to come from institutional investors as opposed to retail investors, is positively associated with the interday measure of crypto-stock return correlation, but this association appears post March 2020. In comparison, although retail attention to monetary policies is positively associated with the intraday measure of crypto-stock return correlation, this positive association appeared well before March 2020 and has not become stronger after.
Our results carry two-sided policy implications. On the one hand, we find little evidence of crypto shocks being transmitted to stock markets, which suggests that price volatility in the crypto markets, however wild, is unlikely to cause systemic risk at the current stage. This finding provides some assurance that turbulence in the crypto market will not result in broader market volatility and economic harm. It also helps address conspiracy theories about recent Bitcoin movements. On the other hand, we observe significant volatility spillovers from stock markets to crypto markets, and such spillovers are highly sensitive to changes in the Fed’s monetary policy. One driving force for such spillovers is the growing presence of institutional investors in crypto, who are sensitive to interest rate changes. In fact, our evidence indicates that the Fed’s policy response to the COVID-19 pandemic led to a structural shift in the crypto-stock correlation by encouraging institutional investors to enter the crypto markets, and this shift appears to persist even after the Fed seeks to reverse course post pandemic. For as long as institutional investors remain interested and involved in the crypto markets, the high crypto-stock correlation is expected to persist as their collective trading, which is sensitive to changes in monetary policies and market liquidity conditions, likely contributes to movements in both markets.
Stephanie Dong is a Ph.D. candidate at the NYU Stern School of Business
Vivian W. Fang is the Honeywell Professor of Accounting at the Carlson School of Management, University of Minnesota.
Wenwei Lin is a Ph.D. candidate at the Carlson School of Management, University of Minnesota.
This post was adapted from their paper, “Tracing Contagion Risk: From Crypto or Stock?” available on SSRN.
