Investor sentiment and green finance indicators: exploring herding behavior in clean versus dirty cryptocurrencies

Abstract

This study examines whether investor sentiment and green finance affect herding in cryptocurrency markets through common or segment specific channels. Using a Quantile on Quantile Regression framework, we show that herding is structurally segmented across Clean and Dirty cryptocurrencies. In the Clean segment, the most coherent channels are Twitter based sentiment and broad sustainability benchmarks, particularly the Twitter Happiness Index and ESG Index, and their effects are associated more with relative repricing than with broad based convergence. In the Dirty segment, the more informative channels are fear based sentiment and tradable green market indicators, especially the Fear and Greed Index, the Global Wind Energy Index, and the Green Bond ETF, which are more closely associated with synchronized trading in states already susceptible to herding. A range of robustness tests preserves this qualitative structure. The findings indicate that sentiment and sustainability effects in cryptocurrency markets are nonlinear, state dependent, and portfolio specific.