Stablecoins and Cryptocurrency Returns : Evidence From Large Bayesian VARs

We study the cross-sectional correlation between the returns on stablecoins and major cryptocurrency pairs. In doing so, we contribute to a growing literature that aims to understand the role of digital assets as an investment. Methodologically, we propose a large-scale Bayesian Vector Autoregressive (BVAR) model which features sparsity in the cross-pair returns correlations as well as the existence of large unexpected returns and time-varying volatility in idiosyncratic risk. Empirically, we focus on the main stablecoin currently traded, namely Tether (USDT), and show that its price changes significantly and positively correlates with future returns on major cryptocurrency pairs, conditional on trading volume. A strategy that exploits the exposure to USDT delivers substantial economic gain out-of-sample relative to an equal-weight market portfolio and a buy-and-hold investment in Bitcoin