Stablecoin Lending Activity as a Predictor of Ethereum Volatility

Summary

This research article examines whether lending activity in USDC, USDT, and DAI helps forecast Ethereum volatility. It describes time-series autocorrelation and partial autocorrelation analysis, then uses Granger causality tests to assess whether past stablecoin activity adds directional predictive information. The article reports repayment frequency as a recurring signal, with USDC associated with next-day volatility and USDT and DAI showing predictive relationships at longer lags. It also describes withdrawals for USDC and borrowed amounts for USDT as potentially informative measures.

The article says it proceeds to compare baseline autoregressive forecasts with models that incorporate stablecoin activity, but the supplied text cuts off before presenting the modeling results. It gives p-values for selected causality findings and interprets timing differences through presumed differences in stablecoin user behavior. Those interpretations are not independently demonstrated in the excerpt, and Granger predictability does not by itself establish that lending activity causes market volatility or produces a profitable trading signal. The analysis is specific to ETH and the described data and methods; the text provides no complete model results or trading evaluation.

Key ideas

• The article tests whether USDC, USDT, and DAI lending activity predicts Ethereum volatility.
• It uses ACF and PACF to describe volatility persistence and guide lag selection.
• Granger tests identify repayment frequency as a predictive measure across the three stablecoins.
• The reported timing differs: USDC is associated with a short lag, while USDT and DAI signals emerge over longer lags.
• The supplied text omits the completed autoregressive model comparison and does not establish profitable trading performance.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.